XIE, Yin et al. ViCToR: Improving Visual Comprehension via Token Reconstruction for Pretraining LMMs. Proceedings of the AAAI Conference on Artificial Intelligence, [S. l.], v. 40, n. 32, p. 27099–27107, 2026. DOI: 10.1609/aaai.v40i32.39924. Disponível em: https://ojs.aaai.org/index.php/AAAI/article/view/39924. Acesso em: 27 may. 2026.